333
Moyosore Imole Student No.: S1626876 m.imole@student.utwente.nl
M.Sc. Research Thesis January 2019
Faculty of Behavioral Management and Social Sciences
Investigating goals to change communication styles: an
empirical study
Supervisor:
prof. dr. Reinout de Vries
Second Supervisor:
Dr. Bas J. Kollöffel
Department of Education Science and Technology University of Twente
P.O. Box 217 7500 AE Enschede The Netherlands
TABLE OF CONTENT
Acknowledgement ... 3
List of Tables ... 4
List of figures ... 4
Summary ... 5
1.0 Introduction ... 7
1.1 Theoretical Conceptual framework ... 10
1.2 Research Questions ... 15
1.3 Research Design ... 20
1.4 Scientific and Practical Relevance ... 21
2.0 Study 1 – Pilot study ... 21
2.1 Methods ... 21
2.1.1 Participants ... 21
2.1.2 Instrument ... 22
2.1.3 Procedure ... 23
2.1.4 Results and discussion ... 24
3.0 Study 2 – Main Study ... 27
3.1 Methods ... 27
3.1.1 Participants ... 27
3.1.2 Instrument ... 28
3.1.3 Procedure ... 29
3.1.4 Results and discussion ... 32
4.0 General discussion ... 40
5.0 Limitations and Future perspectives ... 44
References ... 48
APPENDICES ... 55
Acknowledgement
The journey through my masters programme and this thesis has been a long one. At this time, as I round it off, I would like to express my sincere appreciation to those who helped to make it a rewarding and fulfilling one.
First of all, I would like to appreciate my thesis supervisor Prof. Reinout de Vries for his support right from when I expressed an interest in this topic. His support, valuable insights and professional comments has kept me on the right track to ensure that this thesis was completed.
The discussions and iterations we had on the manuscript has enabled me to produce this work. I also want to thank him for the copious comments on the drafts and the time he took to
painstakingly read through my thesis. I also appreciate my second supervisor Dr. Bas Kolloffel for his insightful comments on my manuscript.
Without naming them all, I would not forget my friends and colleagues at the EST department. I also appreciate my friends in Enschede and elsewhere in The Netherlands and abroad for their constant encouragement to complete this programme.
I cannot forget my parents and siblings for their constant support and encouragement at all times. Thanks to my parents for believing in me. I would like to appreciate my husband who stood by me all through the period of my studies and for helping to take care of the children (Daniel and James). I love you all!
Finally, I would like to thank the Almighty God for sustaining me through this journey.
At last, I can say, this is done and it’s time for the next challenge in my life!
Moyo
Enschede, January 2019
List of Tables
Table 1: Descriptives (Means and Standard deviations) of the social desirability of
communication styles – pilot study...
Table 2: Correlation matrix for background variables (gender, age), communication styles and change goals (CG) Domain level scales - Main study (N = 132) ...
Table 3: One sample t-test results for the comparison to the theoretical mean of the CS and CG ..
Table 4: Mediation regression analysis results when communication satisfaction is used as the mediator between communication styles and change goals ...
Table 5: Mediation regression analysis results when communication effectiveness is used as the mediator between communication styles and change goals ...
Table 6: Mediation regression analysis results when functional maturity is used as the mediator between communication styles and change goals ...
List of figures
Figure 1: Research model – mediation paths ... 14
Title: Investigating Goals to Change Communication Styles: An Empirical Study Summary
Purpose – The goal of the present study was to investigate the relations between people’s communication styles, communication satisfaction, communication effectiveness, functional maturity and their change goals.
Methodology – A pilot study with 11 participants was used to explore the level of social desirability of the six communication style domains namely Expressiveness, Preciseness, Verbal Aggressiveness, Questioningness, Emotionality and Impression Manipulativeness. In the main study, the communication styles and change goals of 132 participants were measured using two questionnaires – the Communication Styles Inventory (CSI) and the Change Goals CSI (C-CSI).
Furthermore, participants’ level of communication satisfaction, communication effectiveness and functional maturity were measured using questionnaires adapted from previous studies.
Regression analysis were performed to test three main hypothesis proposed in the study.
Findings – From the pilot study, Preciseness, Questioningness, Expressiveness and Impression Manipulativeness were identified as a socially desirable (positive) communication style while Verbal Aggressiveness and Emotionality were negative relative to the theoretical scale mean. The results showed that most people wanted to increase with respect to Preciseness and wanted to decrease with respect to Verbal Aggressiveness. CSI-X, CSI-P and CSI-Q showed positive correlations with communication satisfaction, communication effectiveness and functional maturity (except the correlation between CSI-X and communication satisfaction). The other CSI domain scales, namely CSI-VA, CSI-E and CSI-IM had negative correlations with communication satisfaction, communication effectiveness and functional maturity. Multiple regression analysis showed that communication satisfaction mediated the relations between Verbal Aggressiveness
and change goals. Contrary to our expectation however, communication effectiveness and functional maturity did not fully mediate the relation between the other communication styles and the change goals. The findings in this research offer a new glimpse into the relation between people’s communication styles and their change goals and can have implications in personal and leadership training and future communication research.
Keywords: communication styles, change goals, Communication Style Inventory, functional maturity, communication satisfaction, communication effectiveness
1.0 Introduction
Communication – the art of exchanging information, ideas or thoughts – as a generic skill is important in the functioning of students, workers and leaders within the academic environment and also within organizations. This is because, among others, effective communication has been shown to help improve efficiency, foster a good working relationship, boost morale and ensure that set goals are achieved (Albrecht, Johnson, & Walther , 1993; Chao & Ishii, 2003; Hales, 2000;
Maier, Eckert, & Clarkson, 2006). Communication is also an essential part of life, learning and leadership (Awamleh & Gardner, 1999; Towler, 2003; Spangler & House, 1991). From doctor to patient communication (Henbest & Stewart, 1990; Buller & Burgoon, 1996), parent-child communication (Ritchie & Fitzpatrick, 1990) partner communication (Noller & White, 1990), leader communication (Johnson & Belcher, 1998), there is a unanimous agreement in literature as to the importance of effective communication between individuals and groups.
Among students for example, communication takes place in different forms and serves to achieve different objectives. Communication may serve to clarifying statements, deliberate and discuss with peers, develop relationships (McKenna, et al., 2014), elicit viewpoints from other students, offer feedback, mediate conflicts or even to keep a discussion from drifting away.
Communication therefore is seen as an important part of being a student, since being a student would entail the sharing or transmission of information, knowledge, opinions and feelings (McKenna, et al., 2014; De Vries, Bakker-Pieper, Konings, & Schouten, 2013). Besides academic achievement, employers are increasingly placing importance on generic skills needed to compete in the global market (Iksan, et al., 2012). Along with personality, communication is a generic skill that prospective employees may need to succeed in job interviews and in their daily life after
graduation. However, communication is a complex process and different individuals have different ways of communicating. (Yilmaz et al, 2011).
Findings from communication between doctors and patients suggest that a supportive (i.e., friendly and caring) communication style is associated with higher satisfaction among patients (Buller & Buller, 1987; Schmid Mast et al, 2007; De Vries et al, 2010), while a dominant style is linked with less satisfaction among patients and less favourable outcomes, such as malpractice claims (Buller & Burgoon, 1996; Burgoon, et al., 1987). Similarly, in the classroom, results suggest that a supportive communication style is linked to greater satisfaction among students (Prisbell, 1994), while a controlling or dominant style is linked to a less intrinsic motivation (Noels, et al, 1999). In contrast, in family settings, children were found to be more attentive and task-oriented when parents used an ambiguous and dominant communication style (De Vries et al, 2010; Bugental et al, 1999; Rasku Puttonen, 1988).
In order to further understand communication among people, recent studies have increasingly focused on the development of a framework to classify or capture individuals’
communication style. Along this line, the idea of different communication styles has been developed. The proposition behind this is that anything that can be said on the way a person communicates can be encoded or categorized into specific styles. Furthermore, recent studies have shown relationships between communication style and individual personality traits. It has been argued that the main communication style dimensions are sub-components of the two major personality models – the Big Five model (Goldberg, 1990) and the HEXACO model (Ashton &
Lee, 2008). Strong correlations have been found between the scales of the different communication style scales and the personality traits (Leung & Bond, 2001; Weaver, 2005; Eysenck et al, 1985;
De Vries et al, 2013).
In human resource development, a strong relation between communication styles, communication effectiveness and measures of leadership behavior in literature (Neufeld et al, 2010; Klauss & Bass, 1982). Effective communication skills enable leaders to create and disseminate a compelling vision for followers (Bennis & Nanus, 1985; Conger & Kanugo, 1987).
However, leaders are often confronted with different options and communication techniques that are not directly associated with strategic goals (Mayfield & Mayfield, 2002).
In addition, from personality studies, research has shown that for each personality trait, there is a preferred ‘scale position’ in the population and that people tend to express desire to change their personality traits (Hudson & Roberts, 2014; Baumeister, 1994; Kiecolt, 1994). The change goal is generally motivated by discontent with aspect of one’s lives and the feeling that achieving a different personality trait might placate the discontent (Dunlop, Telford, & Morrison, 2012). Furthermore, research on change goals in personality studies show that people generally desire to increase with respect to each of the Big Five personality traits (Hudson & Roberts, 2014).
Emotional stability ranked highest in the desirability order followed by conscientiousness, extraversion, agreeableness and openness. These findings are consistent with other research that show that the Big Five personality dimensions are socially desirable in themselves and can be captured with self-report questionnaires (Dunlop, Telford, & Morrison, 2012).
Even though there is a body of knowledge on the goals to change personality traits, and the links between personality traits and communication style, not much is known about the goals to change communication styles. Furthermore, beside a related study exploring goals to change personality traits (Hudson & Roberts, 2014), to our knowledge no validated method of measuring communication change goals exists. Furthermore, the associated relations between change goals
and communication satisfaction, communication effectiveness and functional maturity is not fully understood.
This research sets out to investigate these open issues. First, using a small-scale pilot study to gain background information, we attempted to understand the social desirability of communication styles within the context of the Communication Styles Inventory (De Vries et al., 2009; De Vries et al., 2013). Building on this, in the main study people’s current communication styles were measured and compared with their change goals. Next, the relations between the communication styles and possible mediators namely communication satisfaction, communication effectiveness, and functional maturity level were investigated. Finally, using a mediation regression model, we investigated whether communication satisfaction, communication effectiveness and/or functional maturity mediate the relation between the identified communication styles and change goals. As a note, it has to be pointed out that the general goal of this research was not to investigate whether people can actually change their communication style but to measure their change goals and the associated relations of change goals to communication satisfaction, communication effectiveness and functional maturity level.
1.1 Theoretical Conceptual framework
In general, communication is the process of exchanging information, from one source (the person giving the information to the receiver (the person to whom information is given) (Iksan, et al., 2012). According to Norton, (Norton R. , 1983), communication style can be defined as “the way one verbally, nonverbally, and paraverbally interacts to signal how literal meaning should be taken, interpreted, filtered, or understood”. Owing to the fact that this definition does not include the interactional aspect of communicative behavior, De Vries et al. (2009) provided a broader definition of communication style as “the characteristic way a person sends verbal, paraverbal, and
nonverbal signals in social interactions denoting (a) who he or she is or wants to (appear to) be, (b) how he or she tends to relate to people with whom he or she interacts, and (c) in what way his or her messages should usually be interpreted.” (De Vries, Bakker-Pieper, Konings, & Schouten, 2013).
Communication scholars have also considered the measurement of the main communication styles as important due to its relevance in different settings in which the transfer of personal and non-personal information, knowledge, feelings, ideas and opinions are important (De Vries et al, 2013). For this purpose, several communication measurement instruments have been developed. Examples of these instruments include the Relational Communication Style (RCS) (Burgoon & Hale, Validation and measurement of the fundamental themes of relational communication, 1987), Communication Style Scale (CSS) (Gundykunst, et al., 1996), and the Communication Style Measure (CSM) (Norton R. , 1978; Norton R. , 1983). These measurement instruments, especially the more recent Communications Style Scale (CSS) have been criticized for the inclusion of scales that focus on interpersonal cognitions and feelings about communication rather than the way communication signals are sent (De Vries et al., 2013). Another issue is with the older instruments (CSM & RCS) is their lack of conformity to psychometric standards, owing to the low reliabilities of the items on the scale. Furthermore, the lack of integration between the models and the absence of an underlying model to define the number and content of the communication styles has also been an issue. To address these drawbacks, a number of researchers have used a lexical approach, as used in personality studies, to uncover the communication styles (Goldberg, 1990; Burgoon, Johnson, & Koch, 1998; Ashton & Lee , 2001). De Vries et al. (2009) performed an empirical study to identify communication style dimensions from adjectives that can be used to describe how a person communicates.
The underlying theory behind the communication style dimensions is that all that is said during the course of a discourse can be encoded in a dictionary. Based on this premise, De Vries et al. (2009) identified seven lexical communication style dimensions, namely Expressiveness, Preciseness, Niceness, Supportiveness, Threateningness, Emotionality and Reflectiveness.
Building on this lexical study, De Vries et al. (2013) reported a new communication styles questionnaire, called the Communication Styles Inventory (CSI). Even though the CSI is similar to the lexical communication styles dimensons in that it captures all the main dimensions of communication identified in the lexical study, a new factor called Verbal Aggressiveness was introduced to account for the three lexical factors Threatening, Niceness and Supportiveness. It was found that it was impossible to construct independent factors that aligned with these lexical factors. Due to this, the CSI represents six behavioral communcation style dimensions namely Expressiveness (extroverted versus withdrawn), Preciseness (expert versus to waffle), Verbal Aggressiveness, Questioningness (named Reflectiveness in the lexical study), Emotionality (stressed, piqued) and Impression Manipulativeness (a measure of deceptive communication style). These six style dimensions form the basis of the present study.
Another factor related to the way people communicate is the functional maturity level.
Functional maturity “refers to the state of human development wherein a person knows who he/she is, knows what he/she wants, and has the courage to pursue it without forgetting to consider others or the bigger picture” (McCuddy & Reeb-Gruber, 2008). The skills associated with maturity in the intellectual, psychological, interpersonal, emotional and moral domains are captured under functional maturity. Overall, eleven skill sets possessed by functionally mature people are identified (McCuddy, Reeb-Gruber, & Thijssen, 2012), namely that functionally mature people (i) are self-aware; (ii) proactively reflect on situations, attitudes, and behaviors, and use those
reflections to guide decisions and actions; (iii) are actively aware of the synergistic relationship between emotion and intellect; (iv) consider alternate perspectives; (v) are aware of the moral implications of their decisions and actions; (vi) balance self-interests with the interests of other individuals, communities, and society at large; (vii) are genuinely willing to risk making mistakes and to learn from them; (viii) know when to let go of disappointment, anger, and/or grudges that get in the way of achieving their goal, that cause them to lose focus, and/or that might significantly change the desired solution or outcome; (ix) can deal effectively with uncertainty; (x) are flexible in switching between behaviors and knowing when a specific behavior is appropriate or inappropriate; and (xi) know how to act in any type of situation, either personally or professionally.
From the foregoing, a functional mature person can be said to possess the ability to communicate as a skilled and effective adult – with the word adult connoting not chronological maturity but functional maturity (Franz, 1998). In this sense, a person’s functional maturity level might be connected to his/her present communication style. Furthermore, functional maturity might also serve as a mediator between the communication styles and the change goals as depicted in the research model (see Figure 1). It may also mean that functional maturity may independently lead to people’s current communication style and change goals, rendering any relationship between communication styles and change goals spurious. If this is true, then communication styles may or may not be related to change goals. For example, an expressive person may be especially likely to desire to increase in expressiveness if expressiveness is a socially desirable communication style. This exact relation between these will be further investigated in this research.
Figure 1: Research model – mediation paths
Besides functional maturity, other mediators that will be explored in this research concern communication satisfaction and communication effectiveness. Investigating people’s satisfaction with the way they presently communicate and people’s opinion of the effectiveness their communication with other people within the context of this research is important because like functional maturity, people’s level of communication satisfaction or effectiveness might help to explain discrepancies between their communication style and their change goals. In other words, like fuctional maturity, communication satisfaction and communication may serve as mediators between people’s communication style and their change goals.
1.2 Research Questions
Based on the foregoing, we first investigated via a simple pilot study the level of social desirability of the communication style scales. Thereafter three fundamental research questions will be answered.
Although the existence of preferable communication styles has not been systematically studied, the existence of preferred or desirable communication styles have been shown within different contexts, namely in assessing counsellor effectiveness (Dauphinais, Dauphinais, &
Rowe, 1981), spousal communication (Hawkins, Weisberg, & Ray, 1980) and culture (Patrick Rau, Li, & Li, 2009). The typical procedure in these studies is to allow participants listen to pre- recorded conversations structured according to different communication styles and rate their level of desirableness. A clear outcome from these studies is the differences in the preference level of participants for different communication styles – from directive versus non-directive/facilitative communication style (Dauphinais, Dauphinais, & Rowe, 1981) and implicit versus explicit communication styles (Patrick Rau, Li, & Li, 2009). Particularly, the non-directive approach was rated as least effective (or least preferred) of the communication styles (Dauphinais, Dauphinais,
& Rowe, 1981).
One limitation of using the pre-recorded conversation is the sensitivity to cultural bias (in the form of accent of the actors employed) as reported in the study (Dauphinais, Dauphinais, &
Rowe, 1981). Due to this, the pilot study is aimed at investigating whether some communication styles are more preferred or more socially desirable than others. This is achieved by using a few selected participants. The idea is to allow participants make a simple rating of their desirability (or preference) level for the items on the communication styles instrument (the CSI) as to whether they find them preferable, non-preferable to neutral. From the responses, the deviations of the
responses from the theoretical mean of the social desirability scale will be investigated.
Furthermore, this will allow us to find out the exact levels of social desirability for each communication styles scale. Due to the small sample size of the pilot study, the inter-rater agreements of the level of social desirability will be investigated. Thereafter, the means from the pilot study will be compared with responses in the main study.
From the main study, three research questions will be answered:
Research question 1:
“What are the relations between people’s existing communication style and their change goals?”
The goal of the first research question is to investigate the connection between people’s identified communication styles and their change goals. Even though one may reasonably expect that some people will not be satisfied with their communication style, whether they want to change it and the reasons and motivations are another question. From personality studies, differences between people’s actual and ideal characteristics have been reported (Higgins, 1987; Markus &
Nurius, 1986). Consequently, the aspiration to reach ideal characteristics could be the trigger to change their communication style. Furthermore, other theorists have argued that the underlying needs of people (e.g. recognition and achievement) (Elliot & Harackiewicz, 1996), life goals (Roberts & Robins, 2000), and personal strivings (Emmons, 1986) motivate the need to change their personality. Since personality has been shown to have strong associations with communication styles (De Vries, et al., 2013), personal aspirations and individual needs may also motivate the need for people to change their communication styles. In investigating people’s change goals, Hudson and Roberts (2014) reported that almost everyone indicated a desire to increase in emotional stability and conscientiousness followed by others who wanted to increase in extraversion, agreeableness and openness, in that order.
Research on change goals in personality traits have shown that people who are low in a certain trait wanted to increase with respect to those traits (Hudson & Roberts, 2014). However, this was attributed to other research that found that the big five personality dimensions are socially desirable in and of themselves (Dunlop, Telford, & Morrison, 2012). In order to answer the question, an attempt will first be made to systematically measure people’s existing communication style within the context of the Communication Style Inventory. Furthermore, an adapted version of the standard Communication Style Inventory, which for the purpose of this research is termed the Change Goals Communication Style Inventory (C-CSI) will be used to evaluate people’s change goals. The results from both instruments will be correlated to investigate the relation between communication styles and change goals.
Hypothesis 1 (H1): People who score low on a socially desirable communication style will express desire to increase with respect to that communication style (implying a negative relation between communication style and change goal).
As an example, if in comparison to the theoretical scale mean of the pilot study, Expressiveness is identified as a socially desirable communication style, people who score low in Expressiveness in the main study will desire to increase with respect to Expressiveness.
The second research question is as follows:
Research Question 2:
“What are the relations between people’s level of communication satisfaction, communication effectiveness, functional maturity and their communication styles?”
Research in personality traits, especially based on the HEXACO model is closely aligned with the Communication Styles Inventory (De Vries et al, 2013). Using the HEXACO model as base, Ashton and Kibeom (2016) investigated age trends in the different personality traits. Honesty
– Humility showed an upward trend of about 1 SD between ages 18 and 60. Emotionality showed a downward trend while Extraversion was upwards as age increased (Ashton & Kibeom, 2016).
On the other hand, Conscientiousness, Openness and Agreeableness showed a upward trend during the teen years while they showed significant differences thereafter by facets (Ashton & Kibeom, 2016).
Furthermore, a look at the relation between personality and communication style as studied by De Vries et al (2013) showed strong negative correlations between Verbal Aggressiveness and Agreeableness while Preciseness showed moderate correlations with Conscientiousness.
Expressiveness showed strong correlation with Extraversion, Questioningness with Openness to experience while Emotionality was found to show strong correlations with HEXACO Emotionality and FFM Neuroticism. On the other hand, Impression Manipulativeness showed strong (negative) correlations with HEXACO Honesty-Humility and FFM Agreeableness.
In related leadership communication studies, relations have been found between leader expressiveness and leader preciseness and various leader criteria (Bakker-Pieper, 2012). These leader criteria were broadly grouped into attitude- and behavior-related on one hand and effectiveness and cognition-related criteria on the other hand. CSI-expressiveness and CSI- preciseness were found to be positively related to communication satisfaction and effectiveness while CSI-verbal aggressiveness was found to show a negative correlation. Expressiveness helps to fulfil the fundamental need to belong while preciseness helps to fulfil the need to reduce uncertainty (Bakker-Pieper, 2012). Satisfaction of these needs has been related to positive individual and social outcomes (DeWall & Bushman, 2011; Hogg & Grieve, 1999). In view of this, we will expect the same relation as in previous studies. Therefore, we hypothesize as follows:
Hypothesis 2 (H2): CSI-expressiveness, CSI-preciseness and CSI-questioningness are positively related to communication satisfaction, communication effectiveness and functional maturity. CSI- verbal aggressiveness, CSI-emotionality and CSI-impression manipulativeness will be negatively related to communication satisfaction, communication effectiveness and functional maturity.
Research question 3:
“Do communication satisfaction, communication effectiveness and functional maturity mediate the relation between people’s communication style and their change goals?”
To show that communication satisfaction, communication effectiveness and functional maturity mediate the relations between communication styles and change goals, we will check whether the relation between communication styles and change goals after controlling for communication satisfaction, communication effectiveness and functional maturity changes from significant to insignificant. This will be done by fulfilling the following conditions (Baron &
Kenny, 1986; Judd & Kenny , 1981):
(i) The causal variable (communication styles) must be correlated with the outcome (change goals) – path 𝑐 in the model (Figure 1)
(ii) The causal variable (communication styles) should correlate with the mediator (communication satisfaction, communication effectiveness and functional maturity) – path 𝑎 in the model
(iii) The mediator should affect the outcome variable – path 𝑏 in the model. In this case, the communication styles and change goals will be used as dependent variables while the change goals will be the outcome
(iv) For complete mediation, the effect of communication styles on change goals while controlling for the mediators should be zero (path 𝑐′ in the model).
If the four conditions are satisfied, then the results will be consistent with the hypothesis that communication satisfaction, communication effectiveness, and/or functional maturity completely mediates the relationship between communication styles and change goals. On the other hand, if the some of the steps are met but not all, then partial mediation is indicated. In other words, if the relationship between communication style and change goal was still significant, but less strong, partial mediation has occurred.
Hypothesis 3 (H3): Communication satisfaction, communication effectiveness and functional maturity mediate the relations between the communication styles and the outcomes of this study, i.e. the change goals.
The model used in this research (see Figure 1) links communication styles to the change goals through the three mediators. This allows us to measure people’s communication style in general, their level of communication satisfaction, communication effectiveness and functional maturity and explore their links to the outcome, their change goals.
1.3 Research Design
Research questions 1-2 were designed as a correlational study to investigate the relations between communication styles and change goals and the relations between communication satisfaction, communication effectiveness and functional maturity and change goals. The third question is expected to broaden the first two research questions by using statistical mediation analysis to investigate whether explanations can be found for people’s change goals by looking at their level of communication satisfaction, communication effectiveness and functional maturity level.
1.4 Scientific and Practical Relevance
In summary, this study is relevant for several reasons. First, it provides a common index for participants to rate their goals to change their communication style. This will unravel the motivations underlying people’s change goals and their concurrent communication styles (De Vries et al, 2013).
Secondly, the study will help to understand the prevalence of the goals to change each communication styles. Based on this, it may become clear whether some communication styles are more socially desirable than others. Finally, on the practical side, being able to identify and measure people’s existing communication style may provide help to people with “undesirable”
styles towards achieving their set goals.
2.0 Study 1 – Pilot study
Study 1 was designed with two goals in mind. First, it served as an initial exploration into the level of social desirability of the different communication style domain and facet level scales. This involved a modification of the scales of the standard CSI to enable the respondents to rate the level of social desirability of the items of the CSI. Second, it was intended to serve as a backdrop for the main study. In other words, once the level of social desirability of the scales are known, this can be used to interpret the results of the main study.
2.1 Methods
2.1.1 Participants
The respondents for the pilot study consisted of 18 invited participants. The criteria for participation in the pilot study was that all respondents should have a minimum of a University
degree and be at least 18 years of age. Participants who did not fully complete the questionnaire and those who did not meet the minimum age and education requirements were excluded. At the end of the exercise, a total number of 11 participants (72% male) were accepted and used in the analysis. Participants age in the accepted sample ranged from 18 to 47 (𝑀 = 35.91, 𝑆𝐷 = 8.64).
Note that even though the small sample size raises questions about the strength of the statistical power for the analysis and the associated risks of strong correlations in the results, the pilot study should be seen as not the main aim of the present investigation but an initial check to prove or disprove the existence of socially desirable communication styles. A detailed study would require the use of larger sample sizes which is beyond the goal of the present research.
2.1.2 Instrument
The measurement instrument for the pilot study was an adapted version of the Communication Style Inventory (CSI) (De Vries et al, 2013). The CSI consists of 96 communication behavior items that are reported in Appendix 1 while the adapted version is reported in Appendix 2. The items are divided equally among the following six domain-level scales (16 items per scale): Expressiveness, Preciseness, Verbal Aggressiveness, Questioningness, Emotionality, and Impression Manipulativeness. Each of the domain-level scales consists of four facets each with four items.
The questionnaires were prepared and administered online with the use of the Qualtrics® platform provided by the University of Twente. The use of digital questionnaire ensures a wide coverage and that data collection and analysis can be simplified. It also provides a clear layout and gives participants the opportunity to complete the questionnaire at their convenience. The questionnaire contained a few demographic questions (age, gender and level of education) followed by instructions asking participants to provide ratings of the each of the 96 item statements
on the CSI as to whether, in their judgment, they consider them desirable or not. Due to the fact that the items in the CSI as reported in De Vries et al (2013) were written in the first person, it was necessary to re-word all the items to the third person in order to prevent the non-personalization of the responses provided. For example, the statement “I always have a lot to say” was written as
“He/she always has a lot to say” (see Appendix 2). The six domain level scales – Expressiveness, Preciseness, Verbal Aggressiveness, Questioningness, Emotionality and Impression Manipulativeness – consist of four facets, each with four items – Expressiveness (talkativeness, conversational dominance, humor, informality), Preciseness (structuredness, thoughtfulness, substantiveness, conciseness), Verbal Aggressiveness (angriness, authoritarianism, derogatoriness, nonsupportiveness), Questioningness (unconventionality, philosophicalness, inquisitiveness, argumentativeness), Emotionality (sentimentality, worrisomeness, tension, defensiveness) and Impression Manipulativeness (ingratiation, charm, inscrutableness, concealingness). Respondents provided ratings on these statements based on a five point Likert- like scale, from 1 (very socially undesirable) to 5 (very socially desirable). Furthermore, it was necessary to split the items into different pages (10 items per page) with the instructions repeated at the top of every page. This was done to prevent unambiguity and ensure that participants had access to the instructions during the course of providing the responses.
2.1.3 Procedure
The first study was a pilot study to determine whether there are desirable (preferred) or undesirable (non-preferred) communication styles. In this study, eleven participants are selected to evaluate the items on the CSI. As minimum selection criteria, participants in this study were required to be at least undergraduates to enable the acquisition of informed and educated responses.
As indicated earlier, the objective of was to use a simple study with high inter-rater reliability to
understand what people think would constitute the ideal/desirable communication profile. The participants were asked to weight each item on CSI as to whether they are desirable or not. Based on the responses provided, an insight into socially desirable (or on the contrary undesirable) communication style was obtained. The questionnaire was administered online using the Qualtrics survey portal, with access granted by the University of Twente. The portal allowed the sharing of the link to the questionnaire via email thereby ensuring efficient data processing. The total time to complete the questionnaire was estimated at 10 minutes.
2.1.4 Results and discussion
As starting point, we investigate the agreement between the responses provided by the 11 raters. This was done by using performing an inter-rater agreement evaluation. The intraclass correlation coefficient (ICC) of all the items of the questionnaire using a two-way mixed model was 0.820 therefore the reliability was considered adequate.
In Table 1, the descriptives (mean and standard deviations) of the social desirability of the CSI domain and facet-level scales obtained in the pilot study are presented. We investigated whether some communication styles are more socially desirable relative to the theoretical mean of the scale used. To do this, a one sample t-test analysis was performed at both the domain and facet level using the theoretical mean scale of 3 as reference. Thus, a combination of the mean difference 𝐷̅ of the domain and facet level scales and their corresponding statistical significance level 𝑝 would show communication styles that are judged as socially desirable or otherwise.
Table 1: Descriptives (Means and Standard deviations) of the social desirability of communication styles – pilot study
Descriptives One sample test (Mean = 3)
𝑀 𝑆𝐷 𝑡 𝑝 𝐷̅ 95% CI
Domains
Expressiveness (X) 3.27 .30 2.928 .015 .267 .06, .47
Preciseness (P) 3.78 .46 5.576 .000 .778 .47, 1.09
Verbal Aggressiveness (VA) 2.18 .69 -4.122 .002 -.856 -1.32, -.39
Questioningness (Q) 3.31 .38 2.703 .022 .307 .05, .56
Emotionality (E) 2.53 .67 -2.315 .043 -.466 -.91, -.02
Impression Manipulativeness (IM) 3.08 .44 .602 .561 .080 -.22, .37 Facets
Talkativeness 3.02 .57 .134 .896 .023 -.36, .40
Conversation dominance 3.34 .74 1.537 .155 .341 -.15, .84
Humor 3.70 .78 3.022 .013 .705 .19, 1.22
Informality 3.00 .49 .000 1.000 .000 -.33, .33
Structuredness 3.66 .68 3.203 .009 .659 .20, 1.12
Thoughtfulness 3.82 .61 4.425 .001 .818 .41, 1.23
Substantiveness 3.68 .36 6.367 .000 .682 .44, .92
Conciseness 3.95 .64 4.943 .001 .955 .52, 1.38
Angriness 2.41 .86 -2.277 .046 -.591 -1.17, -.01
Authoritarianism 2.51 1.16 -1.403 .191 -.492 -1.27, .29
Derogatoriness 2.00 .98 -3.403 .007 -1.000 -1.65, -.35
Nonsupportiveness 1.66 .64 -7.113 .000 -1.341 -1.76, -.92
Unconventionality 2.80 .89 -.766 .462 -.205 -.80, .39
Philosophicalness 3.43 .32 4.503 .001 .432 .22, .65
Inquisitiveness 3.77 .48 5.336 .000 .773 .45, 1.10
Argumentativeness 3.23 .86 .874 .402 .227 -.35, .81
Sentimentality 2.70 .65 -1.507 .163 -.295 -.73, .14
Worrisomeness 2.48 .68 -2.534 .030 -.523 -.98, -.06
Tension 2.27 .93 -2.608 .026 -.727 -1.35, -.11
Defensiveness 2.68 .83 -1.272 .232 -.318 -.88, .24
Ingratiation 3.00 .51 .000 1.000 .000 -.34, .34
Charm 3.16 .875 .603 .560 .159 -.43, .75
Inscrutableness 3.36 .409 2.951 .015 .364 .09, .64
Concealingness 2.80 .640 -1.059 .314 -.205 -.63, .23
Bold: Significant at 𝑝 ≤ 0.05 level, 𝐷̅: mean difference
At the domain level, Preciseness (𝐷̅ = .778; 𝑝 = .000), Questioningness (𝐷̅ = .307; 𝑝 = .022) and Expressiveness (𝐷̅ = .267; 𝑝 = .015) had statistically significant positive mean differences relative to the theoretical scale mean of ‘3’. On the other hand, Verbal Aggressiveness (𝐷̅ = -.856; 𝑝 = .002) and Emotionality (𝐷̅ = -.466; 𝑝 = .043) showed significant negative mean differences. Only Impression Manipulativeness (𝐷̅ = .080; 𝑝 = .561) showed a statistically insignificant mean difference. From this, it appears that majority of respondents viewed Preciseness as a socially desirable style while Verbal Aggressiveness was rated as socially undesirable.
At the facet level, all facet level subsets of Expressiveness (Talkativeness, Conversational Dominance, Humor and Informality) had means higher than the scale midpoint. However, only Humor (𝐷̅ = .705; 𝑝 = .013) was significantly different from the scale mid midpoint. Similarly, all facet level scales under Preciseness (Structuredness, Thoughtfulness, Substantiveness and Conciseness) showed statistically significant positive mean difference relative to the scale midpoint. On the contrary, all facets under Verbal Aggressiveness (Angriness, Authoritarianism, Derogatoriness, Nonsupportiveness) were negative relative to the scale midpoint. Two facet level scales for Questioningness – Philosophicalness (𝐷̅ = .432; 𝑝 = .001) and Inquisitiveness (𝐷̅ = .773; 𝑝 = .000) were significantly different (positive) from the scale midpoint. However, the other two facet level scales – Unconventionality (𝐷̅ = -.205; 𝑝 = .462) and Argumentativeness (𝐷̅ = .227; 𝑝 = .402) – were not significantly different from the scale midpoint. Also, all facets under Emotionality namely Sentimentality, Worrisomeness, Tension and Defensiveness – showed negative mean differences from the scale midpoint. However, while Worrisomeness (𝐷̅ = -.523;
𝑝 = .030) and Tension (𝐷̅ = -.727; 𝑝 = .026) were significantly different from the scale midpoint, Sentimentality (𝐷̅ = -.295; 𝑝 = .163) and Defensiveness (𝐷̅ = -.318; 𝑝 = .232) were not. For the
facet level scales under Impression Manipulativeness, only Inscrutableness (𝐷̅ = .364; 𝑝 = .015) was significantly different from the scale midpoint.
In summary, Expressiveness, Preciseness, Questioningness and Impression Manipulativeness were identified as positive (or socially desirable) communication styles while Verbal Aggressiveness and Emotionality were identified as negative (or socially undesirable) styles.
Consequently, for our first hypothesis, we expect to see a negative correlation between the communication style domains Expressiveness, Preciseness, Questioningness and Impression Manipulativeness and their change goals. In other words, people who score low in these four communication style domains would express desire to increase with respect to them.
3.0 Study 2 – Main Study
3.1 Methods
3.1.1 Participants
The sample size for the main study was determined a priori using the statistical software G*power for a correlation study (t-tests). A sample size of 132 was obtained for an effect size of 0.2 and a statistical power of 0.75. From the invitations sent a total of 175 respondents filled out the questionnaire. Participants who were younger than 18 years old and those who did not complete up to 60 percent of the questions were excluded. This resulted in a total of 132 (37.1% female) valid responses used in further analysis. Mean age of the respondents was (M = 33.05, SD = 7.77).
A significant percentage (85%) of the respondents indicated that they were students. Of this, 66.7%
reported themselves as both students and working.
3.1.2 Instrument
The measurement instrument for the mains study was divided into four main sections, besides the section for the demographics of the participants. The first section was intended to rate the communication styles of participants according to the Communication Style Inventory (CSI) (De Vries et al, 2013), as shown in Appendix 1. In contrast to the pilot study, participants rated their present communication style by answering each item on the CSI item on a five-point Likert scale from “strongly disagree” to “strongly agree”. This allowed participants to indicate to which level the statement fits/aligns with their perception of their current communication style.
In the second section, participants rated their goals to change their communication style used the modified version of the communication style inventory, which is called the Change Goals Communication Style Inventory (C-CSI). The C-CSI, contains the standard 96 items on the CSI.
However, the instructions and the response scales were changed to allow participants to rate how much they would like to change each communication style. For the C-CSI, a seven-point Likert scale from “preferably much less” to “preferably much more”. A seven-point Likert scale was used in order to optimize reliability of the measurements. (Colman, Norris, & Preston, 1997). In this way, participants could indicate whether they wanted to increase, decrease or retain their communication style, with respect to the item in question. Positive and negative scores on these scales represent goals to increase and decrease with respect to the specific item, respectively. The only exceptions to this rule are the 27 negatively worded items in the CSI that had to be re-coded in the post analysis.
A questionnaire based on the development centered paradigm for developing human potential (McCuddy, Reeb-Gruber, & Thijssen, 2012; McCuddy & Reeb-Gruber, 2008) was used to measure people’s functional maturity level. Participants were asked to provide a self-rating of
the ten defining characteristics of a functionally mature person based on a 1-10-point Likert scale.
Thus, the choice of 1 indicates that the item least describes the respondent’s functional maturity level while 10 indicates that the item most describes the respondent’s perception of his/her functional maturity level. In order to personalize the descriptor items on the scale, the original wordings of the defining characteristics were modified, as shown in Appendix 4. For example, the characteristic “functionally mature individuals are self-aware” was written as “I am always aware of myself”.
Another questionnaire was used to measure respondent’s communication satisfaction and communication effectiveness level. The instrument was loosely adapted from previous communication research (Hooijberg, 1996; De Vries, Bakker-Pieper, & Oostenveld, 2010). The instrument contained a combination of nine personalized statements that describes different aspects of communication satisfaction and communication effectiveness, as listed in Appendix 3.
Five of the statements related to the communication satisfaction scale while the other four related to measuring the communication effectiveness scale. Participants provided self-rating of the statements based on a five-point Likert scale from 1 “completely disagree” to 5 “completely agree”. The total completion time for the main study questionnaire was estimated at 25 minutes.
3.1.3 Procedure
A chain-referral (snowball) sampling procedure was used to recruit participants for this research in order to have a broad coverage and representation. Invitations via e-mail were sent out to participants to volunteer to take part in the survey. These participants completed the questionnaire and forwarded the link to the questionnaire to others. The link to the questionnaire was appended to the e-mail invitation. The minimum requirement was that respondents should be at least 18 years old. During the planning phase of the research, some constraints that might impact
on the quality and reliability of the data were identified and mitigating measures were consciously put in place to prevent them. First, it was determined that the success of the study would depend on high response rate and getting a broad representation across different age group, gender, location and work experience. To ensure this, an online questionnaire was chosen as this gave the opportunity to distribute the link by electronic mail to ensure wide coverage. Respondents were also asked to forward the questionnaires to others. Secondly, the timeliness of obtaining the data was another potential constraint identified. To ensure this, reminders were sent to participants on a weekly basis to complete the questionnaire. However, it still took a long time to get all responses in. Thirdly, given the number of questions in the questionnaires, it was anticipated that some participants might find it challenging to complete it in one sitting which might lead to incomplete questionnaires (some completing only the CSI and not the C-CSI or part of both) that might invalidate the purpose of the research. As a mitigation, the possibility of saving the part of the questionnaire completed and coming back at a convenient time to complete the rest was activated.
The responses were collected over a period of seven months before the online portal was deactivated.
To guide against ethical issues with data collection, privacy and data storage, no personal information was collected and participants were invited to indicate if they would like to receive the final report at the end of the research. Furthermore, a clear description of the purpose of the data collection was included in a de-brief section included at the end of the questionnaires. Finally, to assure the quality of the research, the approval of the Ethics Committee of the University of Twente and supervisor was obtained before disseminating the questionnaire. The committee is responsible for ensuring that research by students of the University is performed following the laid-down norms and values of the University.
Table 2: Correlation matrix for background variables (gender, age), communication styles and change goals (CG) Domain level scales - Main study (N = 132)
𝛼 M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1. Gender 1.37 .485 -
2. Age 33.05 7.77 .012 -
3. X .667 2.99 .407 -.159 -.058 -
4. P .811 3.57 .474 -.251 .204⁎ -.098 -
5. VA .795 2.36 .506 -.177⁎ -.225 .282 -.156 -
6. Q .737 3.10 .462 -.311 -.064 .359 .054 .374 -
7. E .879 3.12 .664 .268 .002 -.002 -.313 .001 .155 -
8. IM .777 2.90 .521 -.059 -.322 .230 -.204* .246 .313 .284 -
9. X–CG .329 4.06 .480 -.545 .031 .231* -.101 -.064 .150 .155 .100 -
10. P–CG .799 4.70 .805 -.035 .291 .015 .002 -.322 -.019 .232* -.093 .274 -
11. VA–CG .816 3.16 .839 -.077 -.239 -.020 .019 .349 .154 -.256 .125 -.142 -.599 -
12. Q–CG .670 3.95 .671 -.172 -.070 .124 .065 .127 .372 .199* .246 .305 .059 .097 -
13. E–CG .825 3.53 .856 .006 -.263 -.019 .056 .343 .160 -.085 .215* -.181* -.454 .663 .123 -
14. IM–CG .634 3.61 .684 -.078 -.242 .086 .046 .220* .291 .019 .397 .054 -.381 .450 .613 .487 -
15. CSat. .702 2.50 .713 -.131 .048 -.039 .226 .185⁎ .048 -.440 -.053 -.210⁎ -.466 .450 .043 .378 .283 - 16. CEff. .590 3.48 .542 -.201⁎ .311 .171 .550 -.150 .152 -.306 -.153⁎ .031 .092 -.017 .189 -.120 .037 .24⁎ - 17. FM .881 7.43 1.61 -.144 .344 .062 .348 -.220 .058 -.108 -.066 .116⁎ .278 -.137 .083 -.217⁎ -.113 .006 .318 -
Bold: Correlation is significant at 𝑝 ≤ 0.01 level (2-tailed); Star (*): Correlation is significant at 𝑝 ≤ 0.05 level (2-tailed); CSat: communication satisfaction; CEff: communication effectiveness; FM: functional maturity
